(187e) Investigation of a Counter-Current Crystallization Process for Separation of Systems Forming Solid Solutions

Investigation
of a counter-current crystallization process for separation of systems forming
solid solutions

Erik Temmel¹, Sebastian Wloch¹,
Uwe Müller², Detlef Grawe³, Robert Eilers²,
Heike Lorenz¹, Andreas Seidel-Morgenstern^1,4

¹Max
Planck Institute for Dynamics of Complex Technical Systems, Magdeburg, D-39108,
Germany

²HAPILA GmbH,
Gera, D-07552, Germany

³Jesalis Pharma GmbH, Jena, D- 07745, Germany

⁴Otto
von Guericke University Magdeburg, Chair of Chemical Process Engineering, Magdeburg,
D-39108, Germany

The occurrence of complete
or partial solid solutions in solution crystallization is still the focus of
several recent investigations. One reason is that the underlying mechanisms of
the formation of mixed crystals are not clarified in every detail. On the other
hand this special solid liquid equilibrium (SLE) limits the applicability of commonly
used crystallization processes and complicates separation tasks. Especially for
pharmaceuticals this can lead to a huge economical drawback.

Process
design for crystallization from solution is mainly based on the solid-liquid equilibria (SLE) of the substances of interest [1]. For the,
in case of ternary mixtures, commonly appearing simple eutectic and
intermediate compound-forming systems several opportunities for crystallization
based separation from a solvent are known and are well understood, e.g. processes
based on applying preferential crystallization in case of two dissolved enantiomers.
Besides, the tie lines of systems involving solid solutions do not lead to a
corner of the pure substance in the ternary phase diagram (figure 1). Therefore
it is not possible to separate these mixtures within one crystallization step
and there is no option to apply the commonly used crystallization processes. For
many examples such a miscibility behavior is published including inorganic
systems like salts containing one common ion [1, 2] or simple organic
substances like amino acids [3]. For more complex chiral molecules the
occurrence of complete mixed crystal formation seems to be rare [4]. However,
there are more cases reported where partial solid solutions are present at the
pure enantiomer side of conglomerate and racemic compound-forming chiral systems.

Some
theoretical studies suggest a separation opportunity based on fractional
crystallization [5-7]. By means of repeated dissolution and crystallization it
is possible to separate systems forming solid solutions. This principle can be
realized by exploiting a cascade of batch crystallizers. Furthermore, it was
shown that a counter-current between the solid and liquid phase in such a
process would lead to optimal process performance in terms of yield and
productivity. Nevertheless, realization and automation of a counter-current crystallization
cascade is quite complicated. Between every crystallization stage a solid/liquid
separation is required. Additionally, the automatic transport of the
crystalline solid phase between the stages is difficult and would lead to huge
instrumental effort.    

Figure 1: Ternary phase
diagram of the system K₂SO₄ / (NH₄)₂SO₄
/ H₂O for 30°C and 80°C with conodes
indicated.

Recently an automatic counter-current
crystallization cascade was developed by the HAPILA Gera GmbH [8]. In this
plant exclusively liquid phases are transported. The solid phase is dissolved after
the crystallization to avoid the expensive transport of the crystalline phases
in the solid state. Hence, it is possible to proof this up to now only theoretically
considered separation opportunity for systems involving solid solutions.

In the contribution the SLE
of two different systems forming complete solid solutions will be shown.
Furthermore, the mathematical description of the process of interest will be
explained and verified by a manually operated counter-current crystallization in
a rotary evaporator. The validated model was then used to run a semi-automatic
pilot plant followed by a comparison of theoretical and experimental results.
Finally a partial solid-solution forming system was theoretically investigated in
a simulation study.

References

[1] Mullin, J. W., Crystallization, Boston,
Butterworth-Heinemann (1997)

[2] Dejewska, D.; Szymański,
T., Cryst. Res. Technol., 1998, 33, 757-765

[3] Kurosawa, I. W.; Teja, A. S.; Rousseau, R. W., Fluid Phase Equilib., 2004, 224, 245-249.

[4] Jaques, J.; Collet, André; Wilen,
Enantiomers, racemates, and
resolutions, S. H., Florida, Krieger
(1994)

[5] Lin, S. W; Ng, K. M.; Wibowo,
C., Comput. & Chem. Eng., 2008, 32, 956?970

[6] Balawejder, M.; Galan,
K. ; Elsner, M. P.; Seidel-Morgenstern,
A.; Piatkowski, W.; Antos, D., Chem. Eng. Sci., 2011,
66, 5638?5647

[7] Matz, G., Chem. Ing. Tech.,
1980, 52, 562-570

[8] HAPILA GmbH, Offenlegungsschrift
DE 10 2008 023 833 A1, 2009